*** OPEN LOG/OUTPUT FILE ***

log using "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Output\CROAs.JPART MOVE POLITICIZATION TO PEOS MODELS.06-18-2022.smcl", replace





*** MODELING THE EFFECTS OF CENTRALIZED REPORTING ON THE HANDLING OF EMPLOYEE DISCRIMINATION CASES IN U.S. FEDERAL AGENCIES [KRAUSE & PARK] ***




**** ACCESS DATABASE ***

use "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\CROAs.Krause&Park.06-08-2022.dta", replace





****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************






****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


*** TESTING H1: EVALUATING THE TOTAL NUMBER OF REPORTED DISCRIMINATION CASES [EVENT COUNT OUTCOME EXPONENTIAL MODEL) ****  




****  ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON AGGREGATE COUNTS OF REPORTED CASES OF DISCRIMINATION ///
****  [I.E., # SETTLEMENTS (INFORMAL) + # WITHDRAWN (INFORMAL) + # FORMAL COMPLAINT FILED DISCRIMINATION CASES] ****

** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsumintextreport_count fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. , vce(cluster a_id) aequations
estimates store H1ATE
*
estat endogenous






** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 

eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb , exponential) ///
(direct_reporting lagsumintextreport_count fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. , vce(cluster a_id) atet
estimates store H1ATET
*



 

****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************






*** TESTING H2: EVALUATING THE PROPORTION OF PRIVATE RESOLUTION CASES: TOTAL, SETTLEMENT ONLY, & WITHDRAWN ONLY [FRACTIONAL PROBIT OUTCOME MODEL) ****  





**** ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON PROPORTION OF TOTAL PRIVATE RESOLUTION WITHIN AGENCY ///
**** [I.E., (# SETTLEMENTS + # WITHDRAWN) / (# SETTLEMENTS + # WITHDRAWN + # FORMAL COMPLAINTS FILED) DISCRIMINATION CASES] ****

** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **




** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (intrep_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb , fractional) ///
(direct_reporting lagintrep_prop fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count nonnested) if intrep_prop!=. , vce(cluster a_id) aequations
estimates store H2ATE
* 
estat endogenous







** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (intrep_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb , fractional) ///
(direct_reporting lagintrep_prop fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count nonnested) if intrep_prop!=. , vce(cluster a_id) atet
estimates store H2ATET
*





****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************





**** TESTING H3.A:  ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON PROPORTION OF WITHDRAWN PRIVATE RESOLUTIONS WITHIN AGENCY  ****
**** [I.E., (# WITHDRAWN) / (# SETTLEMENTS + # WITHDRAWN + # FORMAL COMPLAINTS FILED) DISCRIMINATION CASES] ****


** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **




** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, fractional) ///
(direct_reporting lagwithdraw_prop fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count nonnested)  if intrep_prop!=. , vce(cluster a_id) aequations
estimates store H3aATE
*
estat endogenous






** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, fractional) ///
(direct_reporting lagwithdraw_prop fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count nonnested)  if intrep_prop!=. , vce(cluster a_id) atet
estimates store H3aATET
*




 
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************





**** TESTING H3.B: ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON PROPORTION OF SETTLEMENT PRIVATE RESOLUTIONS WITHIN AGENCY  ****
**** [I.E., (# SETTLEMENTS) / (# SETTLEMENTS + # WITHDRAWN + # FORMAL COMPLAINTS FILED) DISCRIMINATION CASES] ****


** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **





** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, fractional) ///
(direct_reporting lagsettle_prop fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count nonnested) if intrep_prop!=. , vce(cluster a_id) aequations
estimates store H3bATE
*
estat endogenous





** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, fractional) ///
(direct_reporting lagsettle_prop fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count nonnested)  if intrep_prop!=. , vce(cluster a_id) atet
estimates store H3bATET
*

coefplot (H1ATE, rename(r1vs0.direct_reporting="ATE") \ H1ATET, rename(r1vs0.direct_reporting="ATET")),ylabel(-50 (100) 350, angle(horizon)) yscale(range (-50 (100) 350)) bylabel(Total Caseloads) vertical yline (0, lcolor(black) lwidth(thin) lpattern(dash)) grid(n) ciopts(recast(rcap) lcolor(dkgreen)) nooffsets msize(medsmall) xlabel("") mcolor(dkgreen) title("Figure SA-2.1" "Total Number of Reported Discrimination", size(medlarge)) saving("FigureSA-21")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-21.gph", replace

coefplot (H2ATE, rename(r1vs0.direct_reporting="H2 ATE") \ H2ATET, rename(r1vs0.direct_reporting="H2 ATET")),ylabel(-0.2 (0.2) 1, angle(horizon)) yscale(range (-0.2 (0.2) 1)) vertical yline (0, lcolor(black) lwidth(thin) lpattern(dash)) grid(n) ciopts(recast(rcap) lcolor(dknavy)) nooffsets msize(medsmall) xlabel("") mcolor(dknavy) title("Figure SA-2.2" "Informal Caseload Rate", size(medlarge)) saving ("FigureSA-22")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-22.gph", replace

coefplot (H3aATE, rename(r1vs0.direct_reporting="H3a ATE") \ H3aATET, rename(r1vs0.direct_reporting="H3a ATET")),ylabel(-0.2 (0.2) 1, angle(horizon)) yscale(range (-0.2 (0.2) 1)) vertical yline (0, lcolor(black) lwidth(thin) lpattern(dash)) grid(n) ciopts(recast(rcap) lcolor(dkorange)) nooffsets msize(medsmall) xlabel("") mcolor(dkorange) title("Figure SA-2.3" "Withdrawn Caseload Rate", size(medlarge)) saving ("FigureSA-23")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-23.gph", replace

coefplot (H3bATE, rename(r1vs0.direct_reporting="H3b ATE") \ H3bATET, rename(r1vs0.direct_reporting="H3b ATET")),ylabel(-0.2 (0.2) 1, angle(horizon)) yscale(range (-0.2 (0.2) 1)) bylabel(Internal Caseloads) vertical yline (0, lcolor(black) lwidth(thin) lpattern(dash)) grid(n) ciopts(recast(rcap) lcolor(cranberry)) nooffsets msize(medsmall) xlabel("") mcolor(cranberry) title("Figure SA-2.4" "Settlement Caseload Rate", size(medlarge)) saving("FigureSA-24")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-24.gph", replace

gr combine FigureSA-21.gph FigureSA-22.gph FigureSA-23.gph FigureSA-24.gph, note("Point Estimates and Corresponding 95% Confidence Intervals", j(right) place(seast) size(vsmall))


****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************





**** SUBTREATMENT MODELS BASED ON LATENT ORGANIZATIONAL FAIRNESS OF ADMINISTRATIVE ENVIRONMENT TERCILES [HIGH, MODERATE, & LOW] ****





****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************






*** TESTING H1: EVALUATING THE TOTAL NUMBER OF REPORTED DISCRIMINATION CASES [EVENT COUNT OUTCOME EXPONENTIAL MODEL) ****  




****  ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON AGGREGATE COUNTS OF REPORTED CASES OF DISCRIMINATION ///
****  [I.E., # SETTLEMENTS (INFORMAL) + # WITHDRAWN (INFORMAL) + # FORMAL COMPLAINT FILED DISCRIMINATION CASES] ****

** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **



****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************



** HIGH ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON UPPER TERCILE [fairnessgsem >= 0.243207] ** 



** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsumintextreport_count  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) aequations
estimates store H_H1ATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsumintextreport_count  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) atet
estimates store H_H1ATET
*



****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** MODERATE ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON INTERTERCILE RANGE [fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207] ***  


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsumintextreport_count fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >=    -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) aequations
estimates store M_H1ATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsumintextreport_count fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) atet
estimates store M_H1ATET
*
 


****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** LOW ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON LOWER TERCILE [fairnessgsem < -0.0520733]; OTHERWISE = 0 ** 
*** NOTE: OVERLAP ASSUMPTION IS VIOLATED REGARDING CROA TREATMENT / NON-CROA TREATMENT ***


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

*eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
*(direct_reporting lagsumintextreport_count fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) aequations
*
*estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
*eteffects (sumintextreport_count fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
*(direct_reporting lagsumintextreport_count fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, *vce(cluster a_id) atet
*
 


 
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************





*** TESTING H2: EVALUATING THE PROPORTION OF PRIVATE RESOLUTION CASES: TOTAL, SETTLEMENT ONLY, & WITHDRAWN ONLY [FRACTIONAL PROBIT OUTCOME MODEL) ****  





**** ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON PROPORTION OF TOTAL PRIVATE RESOLUTION WITHIN AGENCY ///
**** [I.E., (# SETTLEMENTS + # WITHDRAWN) / (# SETTLEMENTS + # WITHDRAWN + # FORMAL COMPLAINTS FILED) DISCRIMINATION CASES] ****

** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **




****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************



** HIGH ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON UPPER TERCILE [fairnessgsem >= 0.243207] ** 



** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (intrep_prop fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagintrep_prop   fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) aequations
estimates store H_H2ATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
eteffects (intrep_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagintrep_prop   fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) atet
estimates store H_H2ATET
*



****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** MODERATE ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON INTERTERCILE RANGE [fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207] ***  


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (intrep_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagintrep_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) aequations
estimates store M_H2ATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (intrep_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagintrep_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) atet
estimates store M_H2ATET
*
 


****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** LOW ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON LOWER TERCILE [fairnessgsem < -0.0520733]; OTHERWISE = 0 ** 


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (intrep_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagintrep_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) aequations
estimates store L_H2ATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (intrep_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagintrep_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) atet
estimates store L_H2ATET
*
 

 






****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************





**** TESTING H3.A:  ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON PROPORTION OF WITHDRAWN PRIVATE RESOLUTIONS WITHIN AGENCY  ****
**** [I.E., (# WITHDRAWN) / (# SETTLEMENTS + # WITHDRAWN + # FORMAL COMPLAINTS FILED) DISCRIMINATION CASES] ****


** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **




****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************



** HIGH ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON UPPER TERCILE [fairnessgsem >= 0.243207] ** 



** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (withdraw_prop fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagwithdraw_prop   fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) aequations
estimates store H_H3aATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagwithdraw_prop   fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) atet
estimates store H_H3aATET
*



****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** MODERATE ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON INTERTERCILE RANGE [fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207] ***  


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagwithdraw_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) aequations
estimates store M_H3aATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagwithdraw_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) atet
estimates store M_H3aATET
*
 


****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** LOW ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON LOWER TERCILE [fairnessgsem < -0.0520733]; OTHERWISE = 0 ** 


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagwithdraw_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) aequations
estimates store L_H3aATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (withdraw_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagwithdraw_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) atet
estimates store L_H3aATET
*
 



****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************





**** TESTING H3.B: ESTIMATE ENDOGENOUS TREATMENT EFFECTS MODELS ON PROPORTION OF SETTLEMENT PRIVATE RESOLUTIONS WITHIN AGENCY  ****
**** [I.E., (# SETTLEMENTS) / (# SETTLEMENTS + # WITHDRAWN + # FORMAL COMPLAINTS FILED) DISCRIMINATION CASES] ****


** NOTE: ALL PRIVATE RESOLUTION CASES OCCUR ONLY AFTER TRADITIONAL COUNSELING OR ALTERNATIVE DISPUTE RESOLUTION COUNSELING HAS BEEN COMPLETED **



** HIGH ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON UPPER TERCILE [fairnessgsem >= 0.243207] ** 



** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (settle_prop fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsettle_prop   fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) aequations
estimates store H_H3bATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsettle_prop   fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= 0.243207, vce(cluster a_id) atet
estimates store H_H3bATET
*



****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** MODERATE ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON INTERTERCILE RANGE [fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207] ***  


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsettle_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) aequations
estimates store M_H3bATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsettle_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem >= -0.0520733 & fairnessgsem < 0.243207, vce(cluster a_id) atet
estimates store M_H3bATET
*
 


****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************


** LOW ORGANIZATIONAL FAIRNESS ADMINISTRATIVE ENVIRONMENT SUB-SAMPLE BASED ON LOWER TERCILE [fairnessgsem < -0.0520733]; OTHERWISE = 0 ** 


** ATE EFFECTS: with ATEs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **

eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsettle_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) aequations
estimates store L_H3bATE
*
estat endogenous



** ATET EFFECTS: with ATETs evaluated as proportion of POM cases [i.e., decentralized reporting cases/control group] **
** NOTE WITHOUT FULL SET OF ESTIMATES SINCE ALL MODEL ESTIMATES ARE IDENTICAL TO ATE REPORTED FULL MODEL **
 
eteffects (settle_prop  fairnessgsem ratio_fsup_msup ratio_minsup_nonmsup lntotworkforce_count politicization_lb, exponential) ///
(direct_reporting lagsettle_prop  fairnessgsem  ratio_fsup_msup ratio_minsup_nonmsup  lntotworkforce_count nonnested) if intrep_prop!=. & fairnessgsem < -0.0520733, vce(cluster a_id) atet
estimates store L_H3bATET
*


coefplot (M_H1ATE, rename(r1vs0.direct_reporting="ATE") \ M_H1ATET, rename(r1vs0.direct_reporting="ATET")),bylabel(Moderate OF) ciopts(recast(rcap) lcolor(dkorange)) mcolor(dkorange) msymbol(circle) || (H_H1ATE, rename(r1vs0.direct_reporting="ATE") \ H_H1ATET, rename(r1vs0.direct_reporting="ATET")),bylabel(High OF) ciopts(recast(rcap) lcolor(green)) mcolor(green) msymbol(circle)||,  nokey norecycle vertical ylabel(-1000 (500) 1000, angle(horizon)) yscale(range (-1000 (500) 1000)) yline (0, lcolor(black) lwidth(thin) lpattern(dash)) nooffsets msize(medsmall) xlabel("") byopts(row(1) note("Low OF: Treatment Overlap Assumption Violated", j(right) place(seast) size(vsmall)) title("Figure SA-3.1" "Total Number of Reported Discrimination", size(med))) saving("FigureSA-31")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-31.gph", replace

coefplot (L_H2ATE, rename(r1vs0.direct_reporting="ATE") \ L_H2ATET, rename(r1vs0.direct_reporting="ATET")), bylabel(Low OF) ciopts(recast(rcap) lcolor(cranberry)) mcolor(cranberry) msymbol(circle)  || (M_H2ATE, rename(r1vs0.direct_reporting="ATE") \ M_H2ATET, rename(r1vs0.direct_reporting="ATET")),bylabel(Moderate OF) ciopts(recast(rcap) lcolor(dkorange)) mcolor(dkorange) msymbol(circle) || (H_H2ATE, rename(r1vs0.direct_reporting="ATE") \ H_H2ATET, rename(r1vs0.direct_reporting="ATET")),bylabel(High OF) ciopts(recast(rcap) lcolor(green)) mcolor(green) msymbol(circle) ||, nokey norecycle vertical ylabel(-0.2 (0.2) 0.8, angle(horizon)) yscale(range (-0.2 (0.2) 0.8)) yline (0, lcolor(black) lwidth(thin) lpattern(dash)) nooffsets msize(medsmall) xlabel("") byopts(row(1) title("Figure SA-3.2" "Informal Caseload Rate", size(med))) saving("FigureSA-32")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-32.gph", replace

coefplot (L_H3aATE, rename(r1vs0.direct_reporting="ATE") \ L_H3aATET, rename(r1vs0.direct_reporting="ATET")), bylabel(Low OF) ciopts(recast(rcap) lcolor(cranberry)) mcolor(cranberry) msymbol(circle) || (M_H3aATE, rename(r1vs0.direct_reporting="ATE") \ M_H3aATET, rename(r1vs0.direct_reporting="ATET")),bylabel(Moderate OF) ciopts(recast(rcap) lcolor(dkorange)) mcolor(dkorange) msymbol(circle) || (H_H3aATE, rename(r1vs0.direct_reporting="ATE") \ H_H3aATET, rename(r1vs0.direct_reporting="ATET")),bylabel(High OF) ciopts(recast(rcap) lcolor(green)) mcolor(green) msymbol(circle) ||, nokey norecycle vertical ylabel(-0.4 (0.4) 1.2, angle(horizon)) yscale(range (-0.4 (0.4) 1.2)) yline (0, lcolor(black) lwidth(thin) lpattern(dash)) nooffsets msize(medsmall) xlabel("") byopts(row(1) title("Figure SA-3.3" "Withdrawn Caseload Rate", size(med))) saving("FigureSA-33")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-33.gph", replace

coefplot (L_H3bATE, rename(r1vs0.direct_reporting="ATE") \ L_H3bATET, rename(r1vs0.direct_reporting="ATET")), bylabel(Low OF) ciopts(recast(rcap) lcolor(cranberry)) mcolor(cranberry) msymbol(circle) || (M_H3bATE, rename(r1vs0.direct_reporting="ATE") \ M_H3bATET, rename(r1vs0.direct_reporting="ATET")),bylabel(Moderate OF) ciopts(recast(rcap) lcolor(dkorange)) mcolor(dkorange) msymbol(circle) || (H_H3bATE, rename(r1vs0.direct_reporting="ATE") \ H_H3bATET, rename(r1vs0.direct_reporting="ATET")),bylabel(High OF) ciopts(recast(rcap) lcolor(green)) mcolor(green) msymbol(circle) ||,  nokey norecycle vertical ylabel(-12 (4) 8, angle(horizon)) yscale(range (-12 (4) 8)) yline (0, lcolor(black) lwidth(thin) lpattern(dash)) nooffsets msize(medsmall) xlabel("") byopts(row(1) title("Figure SA-3.4" "Settlement Caseload Rate", size(med))) saving("FigureSA-34")
graph save "C:\Users\17062\Dropbox\DISCRIMINATION PROJECT (U.S. FEDERAL AGENCIES)\Final Dataset\Graphics\JPART\FigureSA-34.gph", replace

gr combine FigureSA-31.gph FigureSA-32.gph FigureSA-33.gph FigureSA-34.gph, note("Point Estimates and Corresponding 95% Confidence Intervals", j(right) place(seast) size(vsmall))






****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************
****************************************************************************************************************************************************************************




log close

